- Posted on
- Featured Image
A bash-first, practitioner’s guide to fast, reproducible AI on Linux: choose the right distro/kernel and keep firmware current; install dev/Python/monitoring essentials; run workloads in containers (Docker/Podman) with GPU exposure; use pinned venvs; apply performance hygiene (CPU governor, NUMA, fast NVMe, telemetry); monitor with htop/nvtop; includes apt/dnf/zypper commands, NVIDIA container notes, and real-world setups (workstation, headless node, small Slurm cluster).